1. Field of the Invention
This invention concerns fluid flow rate sensing and more particularly the sensing of mass rate of flow of a compressible fluid such as air.
2. Background of the Invention
There is an increasing emphasis on precise fuel management for internal combustion engines due to the need for greatly decreased emissions of undesirable components of the exhaust gases as well as heightened need for improvements in fuel economy. This has led to the development of highly sophisticated fuel management systems which precisely control the delivery of fuel to the engine cylinders in accordance with various operating conditions of the engine. One such factor in determining the proper fuel quantity to be added to the air charge under a given set of operating conditions is the precise rate of mass flow of the air into the engine cylinders.
One approach to determining the air mass flow rate in current usage is to calculate the mass flow rate from the engine intake manifold absolute pressure and measured engine speed and ambient temperatures. However, there is an inherent time lag in such an approach, as well as certain resultant inaccuracies since the point at which the pressure is sensed is remote from the point at which the actual movement of air occurs, i.e., at the throttle plate. The relatively slow response of such an approach imposes certain performance limitations on these fuel control systems, since accuracy is thereby affected.
In automotive applications, cost is of course an all important factor, as is the difficult operating environment, in which wide temperature variations are encountered, as are mechanical vibration and the presence of dirt, water, hydrocarbon vapors. This situation is aggravated by a relatively low level of maintenance.
The traditional approach to measuring mass fluid flow rates has been with Pitot-Venturi Meters, which measure the velocity head of the fluid with temperature and static pressure sensing utilized to convert the dynamic readings to mass flow rates. This approach is deficient in respect to the above applications since the dynamic pressure sensed is rather low at low mass flow rate and thus difficult to accurately measure. The Pitot tube does not sample the entire flow, but rather measures the velocity head only at that point within the passage in which it is located. Since local flow rate variations across the passage are normally encountered, inaccuracy is inherent in this approach. While multiple Pitot tubes have been utilized, and the pressure head distribution across the passageway calculated to account for these variations, inaccuracies still inevitably occur.
Various turbine type flow meters have also been utilized to measure fluid flow. In one such design, a turbine is rotated by the fluid flow and is designed to be of minimum friction and moment of inertia such that the rate of rotation of the turbine is directly proportional to the speed of the fluid. If the fluid is of a compressible nature and/or the system conditions were to change significantly such as to affect the density of the fluid flowing therethrough, fluid pressure and temperature must be measured in conjunction with the rate of rotation to calculate the mass flow rate. This approach would have substantial disadvantages in the context of automotive applications, since the high inertia of the turbine created by its relatively high operating velocities create poor dynamic response since the turbine must be decelerated or accelerated by the changes in fluid flow rates. The different rates of flow rate change encountered in the automotive application described are relatively great and must be sensed with a sensor having a high dynamic response. In addition, while the turbines are designed to be supported for rotation as frictionlessly as possible, the friction that is inevitably present creates hysteresis and loss of accuracy in the sensor. Finally, the delicacy of such sensors would lead to poor field life in the automotive environment.
Several restrained turbine type sensor designs have also heretofore been developed such as disclosed in the U.S. Patents to Kindler et al, No. 2,975,635; Karlby et al, No. 3,164,017; Tatsuya Ichihara et al, No. 3,306,105; and Jennings et al, No. 2,714,310. These patents seek to directly measure mass flow rate by imparting an angular momentum to the field prior to entering a restrained turbine, the angular deflection of the turbine corresponding to the mass rate of flow of the fluid. However, these systems require a rotating turbine to impart the angular momentum to the fluid in addition to the restrained turbine. The accuracy of the system is subject to the limitation that the precise level of angular momentum imparted by the rotating turbine cannot be precisely controlled. In addition, many of these designs do not receive the entire fluid flow within the passage, but rather only sample a portion of the flow which accordingly affects accuracy since nonuniform or locally unpredictable variations in fluid flow rate are usual.
In the Moller U.S. Pat. No. 3,092,994, there is disclosed a system for improving the response of a restrained turbine sensor by a circuit which measures the velocity of change of the turbine element such that it would not require the achievement of the final position before the torque which was being impressed on the turbine could be sensed. However, this approach, as in many of the rotating turbine approaches, require a slider-type contact, adding to the friction of the setup and creating hysteresis. In addition, the turbine has a bearing support for the turbine, adding further to the friction and hysteresis of the unit.
A relatively simple approach to flow measurement disclosed in U.S. Pat. No. 1,665,141 to Mayer describes the use of an bladed screw element which receives all of the fluid flow and the resulting torque being impressed thereon by the change in angular momentum of the fluid as it passes through the bladed screw causing a pivot shaft to be rotated against a spring system with the relative angular deflection of the pivot shaft producing an indication of the torque produced. This system, however, does not lend itself to applications in which widely varying temperature-pressure conditions are to be encountered and also the restriction of the bladed screw creates a relatively high pressure loss in the system. In addition, the pivot shaft is frictionally supported in bearings and involves a relatively great mass such that the dynamic response of the unit would not be high and the hysteresis involved would adversely affect accuracy. The mechanical type readout of course could not be directly utilized in the electronic fuel control system referred to above.
Another mechanical approach is described in the U.S. patent to Meneghelli, No. 2,800,794, in which the elastic couplings are utilized to join a tubing section containing a helical element, with the torque induced by the flow through the helical element causing a relative rotative angular deflection of the tubing sections joined by means of the elastic coupling. The high inertia of the helical element would produce poor dynamic response.
A variation of this system is disclosed in the Kotas U.S. Pat. No. 2,811,855, in which a deflector plate is located in a tubing section. The deflection plate creates a strain in the tube which is sensed by the strain gage arrangement producing electrical signals corresponding to the torque produced by the change in angular momentum of the fluid. While providing an electrical signal output, the sensor according to this patent is subject to external physical abuse since the sensing tubes are externally located and errors in readings could be produced by the stressing of the tube by externally applied forces.
Other approaches which do have suitably high response characteristics include a hot wire anemometer which, while suitable for a laboratory application, requires expensive and elaborate instrumentation such as to preclude its adaptation to the application described.
Accordingly, the prior art as described can be characterized as collectively suffering from a number of deficiencies in the automotive application described: poor dynamic response; excessive hysteresis and friction losses; inaccuracies due to a failure to measure the entire flow, i.e., sampling systems; excessively complex and expensive components; undue delicacy of construction, such as to be unsuitable for the automotive environment; inaccuracies due to a failure to compensate for change in density due to temperature and pressure variations; and excessive pressure losses due to the restrictive effect of the sensor in the flow passage.
It is accordingly an object of the present invention to provide a fluid mass flow rate sensor for sensing the mass rate of flow of a fluid flowing in a passageway which is especially adapted to compressible fluids which are subjected to substantial variations in pressure and temperature in the operating environment.
It is a further object of the present invention to provide such a mass flow rate sensor which has extremely high dynamic rates of response.
It is still a further object of the present invention to provide a mass flow rate sensor which is extremely accurate and which directly measures all of the fluid flow and which minimizes hysteresis and frictional losses.
It is still another object of the present invention to provide such a mass fluid flow rate sensor which is relatively rugged in construction and comprises relatively simple components such as to minimize its cost and its vulnerability to failure in the automotive environment.
It is yet another object of the present invention to provide such fluid flow mass rate sensor which generates an electrical signal which may be directly utilized in electronic fuel control systems.
It is still another object of the present invention to provide such a fluid flow rate sensor which generates electrical signals which vary with mass flow rates in such a way as to be optimally usable in such electronic fuel control systems.
It is still another object of the present invention to provide such a mass fluid flow rate sensor in which the various components are optimized in configuration to produce maximum sensitivity of the sensor.